The number of children being treated with psychostimulant medication for ADHD has been increasing dramatically for a decade. Although also an evidence-based treatment, behavior modification is not as widely recommended. Primary questions are whether medication is needed by all of the children who take it, whether this need could be reduced if behavioral treatments were employed first, and whether the required dosage varies as a function of the presence of and intensity of behavioral treatments. In an ongoing funded NIMH study, we are examining the relative effects of and interactions between different doses of behavioral (none, low, and high) and pharmacological (pl, .15, .3, and .6 mg/kg MPH t.i.d.) treatments for ADHD children by evaluating their separate and combined effects in a controlled summer program (STP). Our preliminary results (1/3 of the Ss) suggest that average total daily dose of MPH can be reduced from 60 mg to 10-20 mg, depending on the concurrent behavioral treatment. Follow-up in the subsequent school year has shown that the need for medication in both the school and home settings is eliminated for many children if behavioral intervention is provided first. Given recent societal concerns about medication use especially in young children and growing concerns about long-term side effects and the absence of beneficial long-term effects, these findings may have great public health importance. The current application seeks to extend the results from our efficacy study to a regular school setting. We have chosen to concentrate on early school years, in a sample of children who are not yet medicated for ADHD, examining whether ongoing behavioral treatment can reduce the need for medication in those children. Children will begin the study assigned to no, low, or high behavior modification conditions in their regular home and school settings. Children's functioning will be assessed weekly, and medication will be added to their treatment regimen when functioning deteriorates to a predefined level. Because treatment for ADHD must be chronic, treatment will continue at the assigned level for a period of 3 years, with a 1-year follow up at the conclusion of treatment. Beyond current functioning, primary dependent measures will be the length of time survived without medication, medication dose as a function of behavioral treatment level, and costs of treatments. Individual differences including child, family, and teacher characteristics will be examined.